AnsweredAssumed Answered

Low Frequency Measurements with AD5933

Question asked by AWilhelm on Jul 20, 2017
Latest reply on Jul 21, 2017 by Snorlax



I am trying to perform low frequency (<10 kHz) measurements using the AD5933. I have run into two main problems. To do so, I understand I need to reduce the master clock of the device. The first problem I experience is that the board begins to become unresponsive at a MCLK frequency at 1 MHz or less. The second problem is that at lower frequencies, measuring a complex impedance returns inaccurate values.


My setup is as follows: I use a function generator attached to the MCLK pin. I use a multiplexer to switch between the calibration impedance and the measurement impedance. The calibration impedance is a 68k Ohm resistor. The Rfb value I use is 5.12 kOhms. 


For the unresponsive problem, the device will randomly hang when the master clock is set to values 1 MHz or less. At lower frequencies, the device is quicker to hang. I believe this is most likely a problem with communication, but haven't been able to figure out how to fix it. I know in some posts, it has been mentioned that the I2C clock needs to be slowed down; however, I don't think this is the problem I am facing as the I2C clock that I am using is 100 kHz, which is relatively much slower than the value of 1 MHz.


For the inaccurate complex impedance measurements, here is an example measurement: At 1Mhz, I calibrated the device and then measured an 81k Ohm resistor. The frequency sweep started at 250 Hz and had a step increment of 1 Hz. The device correctly measured that the impedance of the resistor was 81k Ohms. I then connected the device to a 10 nF capacitor and performed a sweep. The device returned a measured impedance of 95k Ohms, which is different from the expected theoretical value.


Any help with either problem would be much appreciated.